It's easier to understand once you understand how a punnet square works.
Let's let :
A = "Normal" gene
and
a = albino gene
by breeding [albino (aa) X "Normal" (AA] you will have babies with the following genotype : Aa ....all babies will be 100% het for albino because that is the only possible genotype they can have (if the "normal isn't het for anything).
Now, when you breed 2 of those 100% het fot albino babies together ["Normal" het albino (Aa) X "Normal" het for albino(Aa)] , you have the opportunity to produce babies with different genotypes.
Here's the possibilities from a punnet square:
AA = "Normal" gecko het for nothing
Aa = "Normal" gecko het for albino
Aa = "Normal" gecko het for albino
aa = albino gecko
As you can see, the [Aa] genotpye shows up twice in the possible genotypes produced by the punnet square. So, you are twice as likely to produce a [Aa] genotype as you would a completely normal genotype[AA]. So if you produced a normal looking gecko, it is more likely to be het for albino than completely normal, but there is no way of being sure if it is or not......therefore we take the percentage probability that is has for being het.....in this case we use 2/3 and then we label the normal looking babies as 66% het for albino. Only future breeding will either confirm or dismiss that.
Someone please check everything I've written for errors, I wrote this kinda fast . Thanks.
-----
1.0.0 Lampropeltis getula holbrooki
0.1.0 Lampropeltis getula californae
0.0.1 Lampropeltis getula nigra
1.0.0 Lampropeltis mexicana thayeri
2.3.0 Eublapharis macularius macularius
0.0.2 Rhacodactylus ciliatus
0.1.0 Gerrhosaurus major major